CN117252689B - Agricultural user credit decision support method and system based on big data - Google Patents
Agricultural user credit decision support method and system based on big data Download PDFInfo
- Publication number
- CN117252689B CN117252689B CN202311499465.1A CN202311499465A CN117252689B CN 117252689 B CN117252689 B CN 117252689B CN 202311499465 A CN202311499465 A CN 202311499465A CN 117252689 B CN117252689 B CN 117252689B
- Authority
- CN
- China
- Prior art keywords
- semantic
- hard
- data item
- implicit
- training
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 29
- 239000013598 vector Substances 0.000 claims abstract description 130
- 230000004927 fusion Effects 0.000 claims abstract description 103
- 230000003993 interaction Effects 0.000 claims abstract description 27
- 238000012549 training Methods 0.000 claims description 103
- 238000005457 optimization Methods 0.000 claims description 17
- 230000007246 mechanism Effects 0.000 claims description 11
- 238000012502 risk assessment Methods 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 7
- 238000009826 distribution Methods 0.000 claims description 5
- 239000011159 matrix material Substances 0.000 claims description 5
- 238000012937 correction Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 238000003860 storage Methods 0.000 description 16
- 238000011156 evaluation Methods 0.000 description 12
- 238000007726 management method Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 230000008901 benefit Effects 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 230000009471 action Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 3
- 238000013479 data entry Methods 0.000 description 3
- RWSOTUBLDIXVET-UHFFFAOYSA-N Dihydrogen sulfide Chemical compound S RWSOTUBLDIXVET-UHFFFAOYSA-N 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 2
- 238000012271 agricultural production Methods 0.000 description 2
- 238000011157 data evaluation Methods 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 238000012854 evaluation process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000012550 audit Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 230000000153 supplemental effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/09—Supervised learning
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
- Y02A40/10—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in agriculture
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Finance (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Accounting & Taxation (AREA)
- General Business, Economics & Management (AREA)
- Probability & Statistics with Applications (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
A method and a system for supporting credit decision of an agricultural user based on big data relate to the technical field of credit of the agricultural user, and acquire hard condition data and implicit condition data of the agricultural user; respectively carrying out semantic coding on each data item in the hard condition data to obtain a sequence of semantic coding feature vectors of the hard condition data item; respectively carrying out semantic coding on each data item in the implicit condition data to obtain a sequence of semantic coding feature vectors of the implicit condition data item; performing feature interaction fusion on the sequences of the semantic coding feature vectors of the hard condition data items and the semantic coding feature vectors of the latent condition data items to obtain semantic fusion features of the hard-latent data items; based on the semantic fusion characteristics of the hard-implicit data items, determining the default risk level label of the agricultural user, realizing multidimensional assessment of credit risk of the agricultural user, grading the default risk of the agricultural user according to the assessment result, and providing credit decision support for a financial institution.
Description
Technical Field
The present application relates to the field of agricultural user credit technology, and more particularly, to an agricultural user credit decision support method and system based on big data.
Background
Agricultural user credit refers to loans provided by financial institutions to users engaged in agricultural production and management, and is an important support for agricultural development. When the agricultural customer performs credit, the commercial bank or the financial institution evaluates the credit risk of the customer and decides whether to pay or not according to the credit risk evaluation result.
In the assessment process, traditional assessment methods evaluate credit risk of credit customers by constructing a scoring card model with a personal credit scoring system solely through credit reporting. That is, in the evaluation process, the main focus is on the basic credit of the client, i.e. the credit report, but the credit report only can embody the default condition of the client, and cannot embody the income, life conservation, capability and other conditions of the client, so that the traditional evaluation method has single evaluation dimension and no comprehensiveness. Moreover, the evaluation emphasis of the traditional evaluation scheme is on the default risk of the applicant, but neglecting the repayment and consumption capability of the applicant, and the credit risk of the agricultural user is difficult to accurately reflect, so that the insufficient or unreasonable credit supply of the agricultural user by the financial institution is caused.
Accordingly, an agricultural user credit decision support system based on big data is desired.
Disclosure of Invention
The present application has been made in order to solve the above technical problems. The embodiment of the application provides a credit decision support method and system for an agricultural user based on big data, which acquire hard condition data and implicit condition data of the agricultural user; respectively carrying out semantic coding on each data item in the hard condition data to obtain a sequence of semantic coding feature vectors of the hard condition data item; respectively carrying out semantic coding on each data item in the implicit condition data to obtain a sequence of semantic coding feature vectors of the implicit condition data item; performing feature interaction fusion on the sequences of the semantic coding feature vectors of the hard condition data items and the semantic coding feature vectors of the latent condition data items to obtain semantic fusion features of the hard-latent data items; and determining the default risk level label of the agricultural user based on the hard-implicit data item semantic fusion characteristics. In this way, multidimensional assessment of credit risk of the agricultural user can be realized, the agricultural user is classified for default risk according to the assessment result, and credit decision support is provided for a financial institution.
In a first aspect, there is provided a big data based agricultural user credit decision support system comprising:
the agricultural user data acquisition module is used for acquiring hard condition data and implicit condition data of an agricultural user;
the agricultural user hard condition semantic understanding module is used for respectively carrying out semantic coding on each data item in the hard condition data to obtain a sequence of hard condition data item semantic coding feature vectors;
the agricultural user implicit condition semantic understanding module is used for respectively carrying out semantic coding on each data item in the implicit condition data to obtain a sequence of the implicit condition data item semantic coding feature vector;
the agricultural user hard-implicit data item semantic fusion module is used for carrying out feature interaction fusion on the sequence of the hard condition data item semantic coding feature vector and the sequence of the implicit condition data item semantic coding feature vector so as to obtain hard-implicit data item semantic fusion features;
and the agricultural user default risk assessment module is used for determining the default risk level label of the agricultural user based on the rigid-implicit data item semantic fusion characteristics.
In a second aspect, there is provided a big data based agricultural user credit decision support method comprising:
Acquiring hard condition data and implicit condition data of an agricultural user;
respectively carrying out semantic coding on each data item in the hard condition data to obtain a sequence of semantic coding feature vectors of the hard condition data item;
respectively carrying out semantic coding on each data item in the implicit condition data to obtain a sequence of semantic coding feature vectors of the implicit condition data item;
performing feature interaction fusion on the sequence of the semantic coding feature vector of the hard condition data item and the sequence of the semantic coding feature vector of the implicit condition data item to obtain a hard-implicit data item semantic fusion feature;
and determining the default risk level label of the agricultural user based on the hard-implicit data item semantic fusion characteristics.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments or the description of the prior art will be briefly introduced below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a block diagram of a big data based agricultural user credit decision support system in accordance with an embodiment of the present application.
FIG. 2 is a flow chart of a big data based agricultural user credit decision support method according to an embodiment of the present application.
Fig. 3 is a schematic diagram of a big data based agricultural user credit decision support method architecture according to an embodiment of the present application.
Fig. 4 is an application scenario diagram of a big data based agricultural user credit decision support system according to an embodiment of the present application.
Detailed Description
The following description of the technical solutions in the embodiments of the present application will be made with reference to the drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
Unless defined otherwise, all technical and scientific terms used in the examples of this application have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the present application is for the purpose of describing particular embodiments only and is not intended to limit the scope of the present application.
In the description of the embodiments of the present application, unless otherwise indicated and defined, the term "connected" should be construed broadly, and for example, may be an electrical connection, may be a communication between two elements, may be a direct connection, or may be an indirect connection via an intermediary, and it will be understood by those skilled in the art that the specific meaning of the term may be understood according to the specific circumstances.
It should be noted that, the term "first\second\third" in the embodiments of the present application is merely to distinguish similar objects, and does not represent a specific order for the objects, it is to be understood that "first\second\third" may interchange a specific order or sequence where allowed. It is to be understood that the "first\second\third" distinguishing objects may be interchanged where appropriate such that the embodiments of the present application described herein may be implemented in sequences other than those illustrated or described herein.
Having described the basic principles of the present application, various non-limiting embodiments of the present application will now be described in detail with reference to the accompanying drawings.
In one embodiment of the present application, FIG. 1 is a block diagram of a big data based agricultural user credit decision support system according to an embodiment of the present application. As shown in fig. 1, a big data based agricultural user credit decision support system 100 according to an embodiment of the present application includes: the agricultural user data acquisition module 110 is configured to acquire hard condition data and implicit condition data of an agricultural user; the agricultural user hard condition semantic understanding module 120 is configured to perform semantic encoding on each data item in the hard condition data to obtain a sequence of hard condition data item semantic encoding feature vectors; the agricultural user latent condition semantic understanding module 130 is configured to perform semantic encoding on each data item in the latent condition data to obtain a sequence of latent condition data item semantic encoding feature vectors; the agricultural user hard-implicit data item semantic fusion module 140 is configured to perform feature interaction fusion on the sequence of the hard condition data item semantic coding feature vector and the sequence of the implicit condition data item semantic coding feature vector to obtain a hard-implicit data item semantic fusion feature; the agricultural user breach risk assessment module 150 is configured to determine a breach risk level tag of the agricultural user based on the hard-implicit data item semantic fusion feature.
It should be appreciated that credit data for agricultural users typically includes hard condition data and implicit regulatory data, where the hard condition data includes: credit record of agricultural users: including past loan records, repayment capabilities, credit ratings, etc.; income and financial status of agricultural users: including annual revenue, liability statement, cash flow, etc.; the operational history and experience of agricultural users: including operational years, agricultural experience, industry approval, etc.; land ownership and usage rights of agricultural users: related files such as a land certificate, a lease contract and the like are included; market prospect and sales channel of agricultural users: including product market demand, sales channels, partners, etc. Implicit condition data, comprising: experience and expertise of agricultural users: including agricultural experience, skill level, professional training, etc.; management capability and decision capability of agricultural users: including agricultural production management, marketing strategies, risk management, etc.; innovation and adaptation capabilities of agricultural users: including agricultural technology innovation, adaptation to market changes, disaster handling, etc.; interpersonal relationship and collaborative capabilities of agricultural users: including relationships with suppliers, partners, agricultural communities, etc.; market prospect and sales channel of agricultural users: including product market demand, sales channels, partners, etc. In particular, financial institutions may be aided by a combination of these data in assessing the credit status, repayment capacity, business capacity, and market competitiveness of agricultural customers. Hard condition data provides objective financial and business information, while implicit condition data provides more information about the individual quality, ability and potential of agricultural users. Comprehensively considering this data allows for a more accurate assessment of the agricultural user's credit risk and potential returns, thereby making more informed credit decisions.
Based on the above, in the technical scheme of the application, an agricultural user credit decision support system based on big data is provided, which can construct a prediction judgment model by acquiring and analyzing hard condition data and implicit condition data of an agricultural user loan body by utilizing big data technology, realize multidimensional assessment of agricultural user credit risk, classify agricultural users against risk according to assessment results, and provide credit decision support for financial institutions.
Specifically, in the technical scheme of the application, first, hard condition data and implicit condition data of an agricultural user are acquired. It should be appreciated that analysis of the hard and implicit condition data of an agricultural user may assist a financial institution in assessing the credit status, repayment capacity, business capacity, and market competitiveness of the agricultural user. It is worth mentioning that here the hard condition data provides objective financial and business information, while the implicit condition data provides more information about the individual quality, ability and potential of the agricultural user. Comprehensively considering this data may more accurately assess the credit risk and potential return of an agricultural user, thereby helping financial institutions make more intelligent credit decisions.
Then, considering that since the hard adjustment data includes credit records of the agricultural user, income and financial status of the agricultural user, operational history and experience of the agricultural user, land ownership and usage rights of the agricultural user: the method comprises the steps of carrying out semantic understanding and analysis on each data item in hard condition data so as to better represent and express the meaning of each data item, so that the following agricultural user default risk classification task is facilitated, and further carrying out semantic coding on each data item in the hard condition data respectively to obtain a sequence of semantic coding feature vectors of the hard condition data item.
In a specific embodiment of the present application, the agricultural user hard condition semantic understanding module is configured to: and performing vector embedding on each data item in the hard condition data by using a word embedding layer of the hard condition semantic encoder to obtain a sequence of the hard condition data item semantic encoding feature vector.
Next, considering that the implicit condition data includes experience and expertise of the agricultural user, management capability and decision capability of the agricultural user, innovation capability and adaptability of the agricultural user, interpersonal relationship and cooperation capability of the agricultural user, and market prospect and sales channel of the agricultural user, in order to enable semantic understanding and analysis of each data item in the implicit condition data to better represent and express semantic meaning of each data item so as to facilitate a follow-up agricultural user default risk classification task, in the technical scheme of the present application, each data item in the implicit condition data is further subjected to semantic encoding to obtain a sequence of semantic encoding feature vectors of the implicit condition data item.
In a specific embodiment of the present application, the agricultural user implicit conditional semantic understanding module includes: and carrying out vector embedding on each data item in the implicit condition data by using a word embedding layer of the implicit condition semantic encoder to obtain a sequence of the implicit condition data item semantic encoding feature vector.
Further, because hard condition data provides objective financial and business information, implicit condition data provides more information about the individual fitness, capabilities and potential of agricultural users. Such data may help financial institutions assess credit status, repayment capacity, business capacity, and market competitiveness of agricultural users, and comprehensively consider such data to more accurately assess credit risk and potential returns of agricultural users, thereby making more informed credit decisions. Therefore, after the sequence of the hard condition data item semantic coding feature vector and the sequence of the implicit condition data item semantic coding feature vector are obtained, the data item semantic features of the hard condition and the data item semantic features of the implicit condition need to be fused. Based on the above, in the technical scheme of the application, the sequence of the semantic coding feature vector of the hard condition data item and the sequence of the semantic coding feature vector of the implicit condition data item are further fused through a feature interaction fusion unit based on a door mechanism attention model, so that a hard-implicit data item semantic fusion feature vector is obtained, and the expression effect of fusion semantics is improved. In particular, here, the attention model based on the gating mechanism can automatically learn the relevance and importance between different features during feature interaction fusion. By calculating the attention weight, the model can determine the correlation between hard condition data items and implicit condition data items and adjust the weight of the feature according to its degree of contribution to the global semantics. The method has the advantages that more attention can be paid to the features contributing to the fusion semantics, so that the expression capability of the fusion feature vectors is improved, and the expression effect of the hard-implicit data item semantic fusion feature vectors on the fusion semantics is improved.
In a specific embodiment of the present application, the agricultural user hard-implicit data item semantic fusion module is configured to: and the sequence of the semantic coding feature vector of the hard condition data item and the sequence of the semantic coding feature vector of the latent condition data item are subjected to a feature interaction fusion unit based on a door mechanism attention model to obtain a semantic fusion feature vector of the hard-latent data item as the semantic fusion feature of the hard-latent data item.
Further, the agricultural user hard-implicit data item semantic fusion module comprises: the importance degree calculating unit is used for calculating the importance degree of each hard condition data item semantic coding feature vector in the sequence of the hard condition data item semantic coding feature vectors relative to the sequence of the implicit condition data item semantic coding feature vectors so as to obtain a sequence of importance degree vectors; the weight distribution correction unit is used for carrying out weight distribution correction on the sequence of the semantic coding feature vectors of the hard condition data item based on the sequence of the importance degree vector and the sequence of the semantic coding feature vectors of the implicit condition data item so as to obtain a corrected sequence of the semantic coding feature vectors of the hard condition data item; and the semantic interaction fusion unit is used for carrying out attention-based context coding processing on the corrected sequence of the semantic coding feature vectors of the hard condition data item based on the sequence of the semantic coding feature vectors of the implicit condition data item so as to obtain the semantic fusion feature vectors of the hard-implicit data item.
And then, the hard-implicit data item semantic fusion feature vector passes through a classifier to obtain a classification result, wherein the classification result is used for representing the default risk level label of the agricultural user. That is, the classification processing is performed by utilizing the interactive fusion feature information between the hard condition semantic features and the implicit condition semantic features of the agricultural user, so as to realize multi-dimensional assessment of the credit risk of the agricultural user, and the agricultural user is subjected to default risk classification according to the assessment result, so that the support of credit decision is provided for a financial institution.
In one specific embodiment of the present application, the agricultural user breach risk assessment module is configured to: and the hard-implicit data item semantic fusion feature vector is passed through a classifier to obtain a classification result, wherein the classification result is used for representing the default risk level label of the agricultural user.
In one embodiment of the present application, the big data based agricultural user credit decision support system further comprises a training module for training the hard conditional semantic encoder including a word embedding layer, the latent conditional semantic encoder including a word embedding layer, the door mechanism attention model based feature interaction fusion unit, and the classifier. The training module comprises: the training data acquisition unit is used for acquiring training data, wherein the training data comprise training hard condition data and training implicit condition data of an agricultural user, and a true value of a default risk level label of the agricultural user; the training agricultural user hard condition semantic coding unit is used for respectively carrying out semantic coding on each data item in the training hard condition data by using the hard condition semantic coder comprising the word embedding layer so as to obtain a sequence of training hard condition data item semantic coding feature vectors; the training agriculture user latent condition semantic coding unit is used for respectively carrying out semantic coding on each data item in the training latent condition data by using the latent condition semantic coder comprising the word embedding layer so as to obtain a sequence of training latent condition data item semantic coding feature vectors; the training hard regulation-implicit regulation semantic interaction fusion unit is used for enabling the sequence of the training hard condition data item semantic coding feature vector and the sequence of the training implicit condition data item semantic coding feature vector to pass through the door-based mechanical attention model feature interaction fusion unit so as to obtain a training hard-implicit data item semantic fusion feature vector; the feature optimization unit is used for performing regular optimization on the training hard-implicit data item semantic fusion feature vector to obtain an optimized training hard-implicit data item semantic fusion feature vector; the classification loss unit is used for enabling the optimized training hard-implicit data item semantic fusion feature vector to pass through the classifier so as to obtain a classification loss function value; the model training unit is used for training the hard condition semantic encoder comprising the word embedding layer, the latent condition semantic encoder comprising the word embedding layer, the characteristic interaction fusion unit based on the door mechanism attention model and the classifier based on the classification loss function value and propagation in the gradient descending direction.
Wherein, the categorised loss unit is used for: the classifier processes the optimized training hard-implicit data item semantic fusion feature vector with a training classification formula to generate a training classification result, wherein the training classification formula is as follows:wherein->Representing the optimized training hard-implicit data item semantic fusion feature vector,/for>To->Is a weight matrix>To->Representing a bias matrix; and calculating a cross entropy value between the training classification result and a true value as the classification loss function value.
In particular, in the technical solution of the present application, the sequence of the semantic coding feature vectors of the training hard condition data item and the sequence of the semantic coding feature vectors of the training latent condition data item express the semantic of the encoded text of the training hard condition data item and the training latent condition data item respectively, and after the encoded text semantic is passed through a feature interaction fusion unit based on a door mechanism attention model, feature weight adjustment can be performed based on the contribution degree of the encoded text semantic to hidden features, that is, the correlation to global semantics, so that the expression effect of the semantic fusion feature vectors of the training hard-latent data item to the fusion semantics is improved. However, this also results in a large deviation of the expression of the training hard-implicit data item semantic fusion feature vector for the encoded text semantic relative to the semantic dimension of the encoded text, affecting the regression efficiency of the training hard-implicit data item semantic fusion feature vector when it is subjected to classification regression by the classifier.
Therefore, the applicant of the present application performs regular optimization on the training hard-implicit data item semantic fusion feature vector, specifically expressed as: carrying out regular optimization on the training hard-implicit data item semantic fusion feature vector by using the following optimization formula; wherein, the optimization formula is:
wherein,is the +.o. of the training hard-implicit data item semantic fusion feature vector>Characteristic value of individual position->Is the global average value of all feature values of the training hard-implicit data item semantic fusion feature vector, and +.>Is the maximum eigenvalue of the semantic fusion eigenvector of the training hard-implicit data item,/->Is the +.f. of the semantic fusion feature vector of the optimized training hard-implicit data item>Characteristic value of individual position->Representing the value of the natural exponent function as a power of a number.
That is, by applying the concept of the imitative function based on the soft L1 regular expression, the regular expression optimization is based on the parameter vector expression representation of the global distribution of the training hard-implicit data item semantic fusion feature vector to simulate the cost function with the regular expression representation of the classification regression probability, so that the feature manifold representation of the training hard-implicit data item semantic fusion feature vector in the high-dimensional feature space models the point-by-point regression characteristic of the classifier-based weight matrix under the classification regression probability to capture the parameter smooth optimization track of the feature to be classified under the scene geometry of the high-dimensional feature manifold through the parameter space of the classifier model, and the training efficiency of the training hard-implicit data item semantic fusion feature vector under the classification probability regression is improved. In this way, the hard condition data and the implicit condition data of the loan body of the agricultural user can be obtained and analyzed by utilizing the big data technology, so that the multidimensional assessment of the credit risk of the agricultural user is realized, the agricultural user is subjected to default risk classification according to the assessment result, and the support of credit decision is provided for a financial institution.
In one embodiment of the present application, when a client applies for credit, the client can self-hold and fill in information, and can transact credit business through an offer client manager, or transact credit business from a client to a counter, when the client input module detects that new client information is recorded, a data folder named by a client name and an identity card number is established, corresponding client information is stored in the data folder, the data folder is transmitted to a temporary library of the storage module through a wireless signal, and when the client signs up, the corresponding data folder is transferred to the system library for encryption storage.
After the client applies for credit and finishes information input, the data folder to which the client belongs is stored in the temporary library, if the credit subscription is not completed within the specified time of the client, the corresponding data folder in the temporary library is automatically destroyed, so that the confidentiality of the client information is ensured, and meanwhile, the abandoned data in the temporary library can be cleaned, and the storage space of the temporary library is increased.
In this embodiment, the storage module is configured as a data storage center, and may be implemented using one data server or a plurality of data servers.
Wherein the customer subscription time is set according to the credit enterprise specified time.
In the embodiment of the application, the credit data warehouse management system can be realized by a computer with a processor and an information storage machine room, and comprises an input module, a storage module and a reading module, wherein the communication connection is established among the modules through wireless signals, and the signal connection among the modules is established through signal wireless signals, wherein the wireless signals comprise 5G communication, 4G communication, wireless network or local area network communication.
The input module comprises a client input port, a mobile input port and a fixed input port, wherein the client input port is configured as a client active input port, the mobile input port is configured as a client manager mobile intelligent equipment input port, the fixed input port is configured as a counter input port, and the collection of client information is completed through multi-port input.
The storage module is configured as a data storage center, the storage module comprises a plurality of temporary libraries and a system library, the temporary libraries are configured as temporary data storage modules, the system library is configured as business customer information storage modules, the system library data are stored through an encryption communication channel, and the customer information data can be temporarily stored and encrypted stored through the storage module, wherein the encryption storage method can be embedded encryption, and is well known to the person skilled in the art and is not described in detail.
When the client to which the temporary storage data belongs performs credit signing, the data is automatically encrypted and transferred into a system library, and when the client to which the temporary storage data belongs does not sign within the rated time, the data is automatically deleted and cleaned at regular time according to the temporary storage data, so that the data storage pressure of the temporary storage is reduced.
The reading module is configured as a data reading channel, the reading module is respectively connected with the temporary library and the system library in a signal way, and the authority checking module is loaded on the signal channel between the reading module and the system library, wherein the authority checking module is in the prior art and is not described again.
The reading module is installed in a binding mode with the input module.
The input module comprises character information input, image information input and video information input, when the input module creates an input channel, the input module creates an affiliated information storage folder according to the channel input character information, and the input information in the same input channel is assigned to a corresponding folder, and the number of times of client login application ports and the data actively uploaded by the client can be obtained through the reading module, so that the basic data of the client credit application are obtained.
When the reading module and the input module are loaded on the client side at the same time, the corresponding client side can only read the input information of the port in the temporary library, and when the reading module needs to acquire the data in the system library, the user of the user needs to be authorized by the user, so that the confidentiality of the client information is ensured.
The reading module is configured with an active reading module and a passive reading module, the active reading module is configured to actively read the affiliated data by the client, the passive reading module passively acquires affiliated client information for credit risk assessment, and the passive reading module acquires client supplementary submitted data to perfect affiliated credit data.
And when performing credit risk auditing, completing data acquisition according to the client input data and the historical data, wherein the data comprises filling-in data and filling-out selection data, and the filling-in data comprises: name, identification number, credit report, income certificate, bank running water (or electronic bill for payment), liability, optionally filling data including: marital situation, family members, fixed assets, consumption proportion, advanced consumption times, working conditions, social relations and the like, after data are input, credit risk auditing is carried out according to client input data, and decision results are generated according to the credit risk auditing.
In an embodiment of the present application, a credit decision method includes the credit data warehouse management system described above and the following steps:
step 1: and (3) data entry, namely acquiring credit evaluation data to which the customer belongs according to active entry and passive acquisition data of the customer.
Step 2: and (3) data auditing, and evaluating the credit risk of the client according to the credit evaluation data of the client obtained in the step (1).
Step 3: and (3) deciding a result, namely deciding a credit evaluation structure according to the data evaluation auditing structure in the step (2), and finishing the credit evaluation decision of the affiliated client.
Step 1, data entry comprises actively acquiring data and customer supplementary data by a data warehouse, and the data entry comprises necessary filling data and optional filling data.
And 2, data auditing comprises data summarizing, data comparison and threshold judgment, wherein the data summarizing summarizes the input data and the database data obtained in the step 1 to generate affiliated credit auditing data, the data comparison evaluates credit risks according to the content of the credit data, and the credit risk evaluation process is divided into hard condition evaluation and implicit condition evaluation.
And 3, the result decision is to divide the credit data to which the evaluation belongs into one type of data, two types of data and three types of data according to the credit data evaluation result in the step 2, wherein the one type of data is high-quality data, the two types of data are qualified data, and the three types of data are risk data.
In the present application, the step 2 data audit further includes the following steps:
step 21: summarizing the uploading data and the system data to generate credit auditing data;
Step 22: according to the data summarization, credit risk assessment is carried out, and the assessment process is divided into a hard condition and a hidden condition;
step 23: and judging the auditing result according to the set threshold value.
The system data is acquired through a credit data warehouse, and the uploading data is the supplemental data actively uploaded by the client.
The hard conditions include: credit, revenue, fixed assets, job stability (whether it is a high risk job), social rules (i.e., client violations, violation conditions), and the like.
The implicit conditions include: marital status, family status, risk bearing capacity, cultural level, spouse income (married), spouse credit status (married), and the like.
When credit risk assessment is carried out, the hard condition accounts for 60 percent in the assessment, the implicit condition accounts for 40 percent, and the data are divided into one type of data, two types of data and three types of data according to the scoring result, wherein the credit risk of the one type of data is the lowest, the credit risk of the two types of data is high-quality data, the medium risk of the two types of data is qualified data, and the three types of data have credit risk and are risk data.
The score was calculated as follows:
credit risk = hard score60% + recessive Condition->40％
And grading the data of the data according to the final score, finishing grading the data, and deciding whether the data is credit or not.
In summary, the big data-based agricultural user credit decision support system 100 according to the embodiments of the present application is illustrated, which is capable of constructing a prediction judgment model by acquiring and analyzing hard condition data and implicit condition data of an agricultural user loan body by using big data technology, realizing multi-dimensional assessment of agricultural user credit risk, classifying the agricultural user against risk according to the assessment result, and providing credit decision support for a financial institution.
As described above, the big data based agricultural user credit decision support system 100 according to the embodiments of the present application may be implemented in various terminal devices, such as a server or the like for big data based agricultural user credit decision support. In one example, the big data based agricultural user credit decision support system 100 according to embodiments of the present application may be integrated into the terminal device as one software module and/or hardware module. For example, the big data based agricultural user credit decision support system 100 may be a software module in the operating system of the terminal device or may be an application developed for the terminal device; of course, the big data based agricultural user credit decision support system 100 could equally be one of the numerous hardware modules of the terminal device.
Alternatively, in another example, the big data based agricultural user credit decision support system 100 and the terminal device may also be separate devices, and the big data based agricultural user credit decision support system 100 may be connected to the terminal device through a wired and/or wireless network and communicate the interaction information in a agreed data format.
In one embodiment of the present application, FIG. 2 is a flow chart of a big data based agricultural user credit decision support method according to an embodiment of the present application. Fig. 3 is a schematic diagram of a big data based agricultural user credit decision support method architecture according to an embodiment of the present application. As shown in fig. 2 and 3, the big data-based agricultural user credit decision support method includes: 210, acquiring hard condition data and implicit condition data of an agricultural user; 220, respectively carrying out semantic coding on each data item in the hard condition data to obtain a sequence of semantic coding feature vectors of the hard condition data item; 230, respectively carrying out semantic coding on each data item in the implicit condition data to obtain a sequence of semantic coding feature vectors of the implicit condition data item; 240, performing feature interaction fusion on the sequence of the semantic coding feature vectors of the hard condition data item and the sequence of the semantic coding feature vectors of the implicit condition data item to obtain a hard-implicit data item semantic fusion feature; and 250, determining the default risk level label of the agricultural user based on the hard-implicit data item semantic fusion characteristics.
It will be appreciated by those skilled in the art that the specific operation of the various steps in the above-described big data based agricultural user credit decision support method has been described in detail above with reference to the description of the big data based agricultural user credit decision support system of fig. 1, and thus, duplicate descriptions thereof will be omitted.
Fig. 4 is an application scenario diagram of a big data based agricultural user credit decision support system according to an embodiment of the present application. As shown in fig. 4, in the application scenario, first, hard condition data (e.g., C1 as illustrated in fig. 4) and implicit condition data (e.g., C2 as illustrated in fig. 4) of an agricultural user are acquired; the obtained hard condition data and implicit condition data are then input into a server (e.g., S as illustrated in fig. 4) deployed with a big data based agricultural user credit decision support algorithm, wherein the server is capable of processing the hard condition data and the implicit condition data based on the big data agricultural user credit decision support algorithm to determine an agricultural user' S breach risk level tag.
The basic principles of the present application have been described above in connection with specific embodiments, however, it should be noted that the advantages, benefits, effects, etc. mentioned in the present application are merely examples and not limiting, and these advantages, benefits, effects, etc. are not to be considered as necessarily possessed by the various embodiments of the present application. Furthermore, the specific details disclosed herein are for purposes of illustration and understanding only, and are not intended to be limiting, as the application is not intended to be limited to the details disclosed herein as such.
The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Finally, it is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or terminal device comprising the element.
The foregoing description has been presented for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of the application to the form disclosed herein. Although a number of example aspects and embodiments have been discussed above, a person of ordinary skill in the art will recognize certain variations, modifications, alterations, additions, and subcombinations thereof.
Claims (8)
1. An agricultural user credit decision support system based on big data, comprising:
the agricultural user data acquisition module is used for acquiring hard condition data and implicit condition data of an agricultural user;
the agricultural user hard condition semantic understanding module is used for respectively carrying out semantic coding on each data item in the hard condition data to obtain a sequence of hard condition data item semantic coding feature vectors;
the agricultural user implicit condition semantic understanding module is used for respectively carrying out semantic coding on each data item in the implicit condition data to obtain a sequence of the implicit condition data item semantic coding feature vector;
the agricultural user hard-implicit data item semantic fusion module is used for carrying out feature interaction fusion on the sequence of the hard condition data item semantic coding feature vector and the sequence of the implicit condition data item semantic coding feature vector so as to obtain hard-implicit data item semantic fusion features;
The agricultural user default risk assessment module is used for determining a default risk level label of the agricultural user based on the rigid-implicit data item semantic fusion characteristics;
the system also comprises a training module for training the hard condition semantic encoder comprising the word embedding layer, the latent condition semantic encoder comprising the word embedding layer, the characteristic interaction fusion unit based on the door mechanism attention model and the classifier;
the training module comprises:
the training data acquisition unit is used for acquiring training data, wherein the training data comprise training hard condition data and training implicit condition data of an agricultural user, and a true value of a default risk level label of the agricultural user;
the training agricultural user hard condition semantic coding unit is used for respectively carrying out semantic coding on each data item in the training hard condition data by using the hard condition semantic coder comprising the word embedding layer so as to obtain a sequence of training hard condition data item semantic coding feature vectors;
the training agriculture user latent condition semantic coding unit is used for respectively carrying out semantic coding on each data item in the training latent condition data by using the latent condition semantic coder comprising the word embedding layer so as to obtain a sequence of training latent condition data item semantic coding feature vectors;
The training hard regulation-implicit regulation semantic interaction fusion unit is used for enabling the sequence of the training hard condition data item semantic coding feature vector and the sequence of the training implicit condition data item semantic coding feature vector to pass through the door-based mechanical attention model feature interaction fusion unit so as to obtain a training hard-implicit data item semantic fusion feature vector;
the feature optimization unit is used for performing regular optimization on the training hard-implicit data item semantic fusion feature vector to obtain an optimized training hard-implicit data item semantic fusion feature vector;
the classification loss unit is used for enabling the optimized training hard-implicit data item semantic fusion feature vector to pass through the classifier so as to obtain a classification loss function value;
the model training unit is used for training the hard condition semantic encoder comprising the word embedding layer, the latent condition semantic encoder comprising the word embedding layer, the characteristic interaction fusion unit based on the door mechanism attention model and the classifier based on the classification loss function value and through gradient descent direction propagation;
carrying out regular optimization on the training hard-implicit data item semantic fusion feature vector by using the following optimization formula; wherein, the optimization formula is: Wherein (1)>Is the +.o. of the training hard-implicit data item semantic fusion feature vector>Characteristic value of individual position->Is what is shown asThe training hard-recessive data item semantically fuses the global average value of all feature values of the feature vector, and +.>Is the maximum eigenvalue of the semantic fusion eigenvector of the training hard-implicit data item,/->Is the +.f. of the semantic fusion feature vector of the optimized training hard-implicit data item>Characteristic value of individual position->Representing the value of the natural exponent function as a power of a number.
2. The big data based agricultural user credit decision support system of claim 1, wherein the agricultural user hard condition semantic understanding module is configured to:
and performing vector embedding on each data item in the hard condition data by using a word embedding layer of the hard condition semantic encoder to obtain a sequence of the hard condition data item semantic encoding feature vector.
3. The big data based agricultural user credit decision support system of claim 2, wherein the agricultural user implicit conditional semantic understanding module comprises:
and carrying out vector embedding on each data item in the implicit condition data by using a word embedding layer of the implicit condition semantic encoder to obtain a sequence of the implicit condition data item semantic encoding feature vector.
4. The big data based agricultural user credit decision support system of claim 3, wherein the agricultural user hard-implicit data item semantic fusion module is configured to: and the sequence of the semantic coding feature vector of the hard condition data item and the sequence of the semantic coding feature vector of the latent condition data item are subjected to a feature interaction fusion unit based on a door mechanism attention model to obtain a semantic fusion feature vector of the hard-latent data item as the semantic fusion feature of the hard-latent data item.
5. The big data based agricultural user credit decision support system of claim 4, wherein the agricultural user hard-implicit data item semantic fusion module comprises:
the importance degree calculating unit is used for calculating the importance degree of each hard condition data item semantic coding feature vector in the sequence of the hard condition data item semantic coding feature vectors relative to the sequence of the implicit condition data item semantic coding feature vectors so as to obtain a sequence of importance degree vectors;
the weight distribution correction unit is used for carrying out weight distribution correction on the sequence of the semantic coding feature vectors of the hard condition data item based on the sequence of the importance degree vector and the sequence of the semantic coding feature vectors of the implicit condition data item so as to obtain a corrected sequence of the semantic coding feature vectors of the hard condition data item;
And the semantic interaction fusion unit is used for carrying out attention-based context coding processing on the corrected sequence of the semantic coding feature vectors of the hard condition data item based on the sequence of the semantic coding feature vectors of the implicit condition data item so as to obtain the semantic fusion feature vectors of the hard-implicit data item.
6. The big data based agricultural user credit decision support system of claim 5, wherein the agricultural user breach risk assessment module is configured to: and the hard-implicit data item semantic fusion feature vector is passed through a classifier to obtain a classification result, wherein the classification result is used for representing the default risk level label of the agricultural user.
7. The big data based agricultural user credit decision support system of claim 6, wherein the categorical loss unit is configured to:
the classifier processes the optimized training hard-implicit data item semantic fusion feature vector with a training classification formula to generate a training classification result, wherein the training classification formula is as follows:
wherein->Representing the optimized training hard-implicit data item semantic fusion feature vector,/for >To->Is a weight matrix>To->Representing a bias matrix;
and calculating a cross entropy value between the training classification result and a true value as the classification loss function value.
8. An agricultural user credit decision support method based on big data, comprising:
acquiring hard condition data and implicit condition data of an agricultural user;
respectively carrying out semantic coding on each data item in the hard condition data to obtain a sequence of semantic coding feature vectors of the hard condition data item;
respectively carrying out semantic coding on each data item in the implicit condition data to obtain a sequence of semantic coding feature vectors of the implicit condition data item;
performing feature interaction fusion on the sequence of the semantic coding feature vector of the hard condition data item and the sequence of the semantic coding feature vector of the implicit condition data item to obtain a hard-implicit data item semantic fusion feature;
determining a default risk level label of the agricultural user based on the hard-implicit data item semantic fusion characteristics;
the method further comprises the steps of: training the hard condition semantic encoder comprising the word embedding layer, the latent condition semantic encoder comprising the word embedding layer, the characteristic interaction fusion unit based on the door mechanism attention model and the classifier;
The training step comprises the following steps:
acquiring training data, wherein the training data comprises training hard condition data and training implicit condition data of an agricultural user, and a true value of a default risk level label of the agricultural user;
respectively carrying out semantic coding on each data item in the training hard condition data by using the hard condition semantic coder containing the word embedding layer to obtain a sequence of semantic coding feature vectors of the training hard condition data item;
respectively carrying out semantic coding on each data item in the training implicit condition data by using the implicit condition semantic coder containing the word embedding layer to obtain a sequence of training implicit condition data item semantic coding feature vectors;
the training hard condition data item semantic coding feature vector sequence and the training implicit condition data item semantic coding feature vector sequence pass through the door mechanism attention model-based feature interaction fusion unit to obtain a training hard-implicit data item semantic fusion feature vector;
performing regular optimization on the training hard-implicit data item semantic fusion feature vector to obtain an optimized training hard-implicit data item semantic fusion feature vector;
The optimized training hard-implicit data item semantic fusion feature vector passes through the classifier to obtain a classification loss function value;
training the hard conditional semantic encoder comprising the word embedding layer, the latent conditional semantic encoder comprising the word embedding layer, the door-based mechanical attention model feature interaction fusion unit and the classifier based on the classification loss function value and through gradient descent direction propagation;
carrying out regular optimization on the training hard-implicit data item semantic fusion feature vector by using the following optimization formula; wherein, the optimization formula is:
wherein (1)>Is the +.o. of the training hard-implicit data item semantic fusion feature vector>Characteristic value of individual position->Is the global average value of all feature values of the training hard-implicit data item semantic fusion feature vector, and +.>Is the maximum eigenvalue of the semantic fusion eigenvector of the training hard-implicit data item,/->Is the +.f. of the semantic fusion feature vector of the optimized training hard-implicit data item>Characteristic value of individual position->Representing the value of the natural exponent function as a power of a number.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311499465.1A CN117252689B (en) | 2023-11-13 | 2023-11-13 | Agricultural user credit decision support method and system based on big data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311499465.1A CN117252689B (en) | 2023-11-13 | 2023-11-13 | Agricultural user credit decision support method and system based on big data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117252689A CN117252689A (en) | 2023-12-19 |
| CN117252689B true CN117252689B (en) | 2024-02-13 |
Family
ID=89133500
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311499465.1A Active CN117252689B (en) | 2023-11-13 | 2023-11-13 | Agricultural user credit decision support method and system based on big data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117252689B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117670527B (en) * | 2024-02-02 | 2024-04-09 | 北京佳格天地科技有限公司 | Method and system for determining peasant household loan credit limit based on land parcel data |
| CN117710096B (en) * | 2024-02-05 | 2024-04-16 | 北京佳格天地科技有限公司 | Post-loan risk monitoring method and system for agricultural loan based on land parcel information |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113361259A (en) * | 2021-06-04 | 2021-09-07 | 浙江工业大学 | Service flow extraction method |
| CN115293910A (en) * | 2022-08-04 | 2022-11-04 | 杭州衡泰技术股份有限公司 | Intelligent enterprise cash flow rating system based on financial big data |
| CN115983972A (en) * | 2022-09-21 | 2023-04-18 | 上海磐农信息科技有限公司 | Construction method of agricultural field wind control model |
| CN116308754A (en) * | 2023-03-22 | 2023-06-23 | 广州信瑞泰信息科技有限公司 | Bank credit risk early warning system and method thereof |
| CN116564355A (en) * | 2023-05-29 | 2023-08-08 | 西安电子科技大学 | Multi-mode emotion recognition method, system, equipment and medium based on self-attention mechanism fusion |
| CN116777608A (en) * | 2023-08-25 | 2023-09-19 | 北京佳格天地科技有限公司 | Agricultural financial risk supervision system, method and storage medium based on big data |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8799150B2 (en) * | 2009-09-30 | 2014-08-05 | Scorelogix Llc | System and method for predicting consumer credit risk using income risk based credit score |
-
2023
- 2023-11-13 CN CN202311499465.1A patent/CN117252689B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113361259A (en) * | 2021-06-04 | 2021-09-07 | 浙江工业大学 | Service flow extraction method |
| CN115293910A (en) * | 2022-08-04 | 2022-11-04 | 杭州衡泰技术股份有限公司 | Intelligent enterprise cash flow rating system based on financial big data |
| CN115983972A (en) * | 2022-09-21 | 2023-04-18 | 上海磐农信息科技有限公司 | Construction method of agricultural field wind control model |
| CN116308754A (en) * | 2023-03-22 | 2023-06-23 | 广州信瑞泰信息科技有限公司 | Bank credit risk early warning system and method thereof |
| CN116564355A (en) * | 2023-05-29 | 2023-08-08 | 西安电子科技大学 | Multi-mode emotion recognition method, system, equipment and medium based on self-attention mechanism fusion |
| CN116777608A (en) * | 2023-08-25 | 2023-09-19 | 北京佳格天地科技有限公司 | Agricultural financial risk supervision system, method and storage medium based on big data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117252689A (en) | 2023-12-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117252689B (en) | Agricultural user credit decision support method and system based on big data | |
| Zhang et al. | A credit risk assessment model based on SVM for small and medium enterprises in supply chain finance | |
| Paleologo et al. | Subagging for credit scoring models | |
| US20150026035A1 (en) | Data analytics model for loan treatment | |
| EP2506207A1 (en) | Computer system, computer-implemented method and computer program product for managing operations of a system | |
| US7970640B2 (en) | Purchasing optimization system | |
| WO2020035075A1 (en) | Method and system for carrying out maching learning under data privacy protection | |
| Liu et al. | Can listing information indicate borrower credit risk in online peer-to-peer lending? | |
| US10984446B1 (en) | Method and system for predicting relevant offerings for users of data management systems using machine learning processes | |
| Alles et al. | Reporting 4.0: Business reporting for the age of mass customization | |
| Ashofteh et al. | A non-parametric-based computationally efficient approach for credit scoring | |
| Ngoc Hung et al. | Factors affecting the quality of financial statements from an audit point of view: A machine learning approach | |
| Hasheminejad et al. | Data mining techniques for analyzing bank customers: A survey | |
| US20230186385A1 (en) | Computer-implemented system and method of facilitating artificial intelligence based lending strategies and business revenue management | |
| Jurgovsky | Context-aware credit card fraud detection | |
| Pinheiro et al. | Introduction to Statistical and Machine Learning Methods for Data Science | |
| Fu et al. | An Alternative Credit Scoring System in China's Consumer Lending Market: A System Based on Digital Footprint Data | |
| Yan et al. | A Deep Machine Learning‐Based Assistive Decision System for Intelligent Load Allocation under Unknown Credit Status | |
| Yu | [Retracted] Big Data Analytics and Discrete Choice Model for Enterprise Credit Risk Early Warning Algorithm | |
| Liu et al. | Unobservable risks in mortgage contract choice | |
| CN112862602B (en) | User request determining method, storage medium and electronic device | |
| CN117372132B (en) | User credit score generation method, device, computer equipment and storage medium | |
| CN116993490B (en) | Automatic bank scene processing method and system based on artificial intelligence | |
| Dai et al. | Financial Risk Early‐Warning of Neusoft Group Based on Support Vector Machine | |
| US20240135402A1 (en) | Systems, methods, and devices for automatic dataset valuation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |